The Challenge
Across this transport operator’s enterprise - from operations and maintenance to scheduling and customer service - the business relied on dozens of defined processes and SOPs embedded in documents, SharePoint folders, slide decks, and process repositories.
Many of these processes were created years ago by external consultants, analysts, or transformation teams - people no longer with the organisation.
While some SOPs were regulated and needed static anchors, most processes in daily operations required adaptability: conditions changed constantly, whether due to customer needs, staff availability, system constraints, or environment triggers.
Yet, the reality was that the organisation was running a live, dynamic business on dead process documents.
This created widespread operational drag and growing risk:
- No one knew what the ‘live’ process was - staff were following instincts, not up-to-date logic
- Process drift occurred between locations, shifts, and teams - eroding standardisation and control
- Exceptions were handled differently every time, increasing regulatory exposure
- Training new hires meant shadowing, not learning - because documentation couldn’t keep up
- Critical tasks were forgotten, delayed, or handled inconsistently because no live system was guiding execution
- Processes weren’t just under-optimised - they were invisible and unmanaged in real time.
- This wasn’t a process problem. It was a process execution failure.
LPx (Living Process Execution) was deployed across the most critical operations - replacing static SOPs and theoretical flowcharts with a live execution environment.
Here’s how it worked:
- LPx captured the core structure of regulated and operational processes
- But instead of freezing them, it connected them to real-time data and events (e.g., demand changes, delays, staff availability, service alerts)
- Processes adjusted dynamically while still adhering to compliance thresholds
Custom GPTs were layered in to support this fluidity:
- Staff could ask, “What do I do next if the service is cancelled?” - and get a context-aware answer
- GPT also acted as a training co-pilot - helping staff learn the why, not just the what
Meanwhile, intelligent nudges and reminders were embedded via LPx triggers:
- “You haven’t closed the disruption record from this morning - would you like to do it now?”
- “Based on the current status, it’s time to initiate the asset handoff process - click here to proceed.”
- LPx turned theoretical processes into a live guidance and accountability engine.
The results
- 74% increase in real-time process adherence across all ops teams
- 53% reduction in time-to-completion for core workflows (fewer missed handoffs, fewer rework loops)
- 60% drop in staff-reported confusion about what to do next
- Onboarding time reduced by 48%, thanks to GPT-led process guidance
- Improved compliance posture: audits showed a 2x increase in traceable, policy-aligned execution
- Processes updated continuously, with logic adapted in response to events — not locked in quarterly reviews